The paper proposes compressed delayed-information projection (CDIP), which handles delayed-arrival acoustic position measurements in a 15-dimensional error-state Kalman filter (ESKF), retaining source-associated IMU snapshots and the current-to-snapshot cross-covariances, so delayed corrections can be projected without repeating the filter from the start.
The authors propose the TacO framework to systematically evaluate tactile sensors based on manipulation policy performance, covering four modalities: visual, acoustic, magnetic, and resistive. They train separate policies for three tasks—pick-and-place with unknown mass, object reorientation, and plug insertion—and analyze how sensor properties such as spatial resolution, shear sensing, tactile representation, and material friction affect task performance. The results show that the value of tactile information depends on the sensor modality, material properties, and specific manipulation task, rather than being universally beneficial; the sensors, code, data, and hardware designs will be made public on the project website.
The authors propose AlphaAdj, a vision-to-control navigation framework that uses first-person RGB observations and a vision-language model (VLM) to adjust the conservativeness of control barrier functions (CBFs) according to visual semantic context. The VLM outputs a scalar contextual caution signal that dynamically modulates the CBF parameter alpha, while a geometry- and speed-dependent cap bounds how aggressive the behavior can be. To cope with VLM latency, inference and the control loop run asynchronously, and the semantic signal is taken directly from token-level output probabilities rather than from full autoregressive text generation.
FlowTouch is a view-invariant visuo-tactile prediction model that uses an object's local 3D mesh to encode information for predicting tactile patterns, enabling tactile feedback even before the object is touched. The method combines scene reconstruction with a Flow Matching-based image generation model. Results show that it narrows the sim-to-real gap and generalizes to new sensor instances, and the generated tactile images can also be used for downstream grasp stability prediction. Code, dataset, and videos are publicly available.
WildOS is a unified system for long-range, open-vocabulary object search that combines safe geometric exploration with semantic visual reasoning for complex outdoor environments without prior maps and with limited depth sensing. The system builds a sparse navigation graph to maintain spatial memory and uses a foundation-model-based vision module, ExploRFM, to score the graph's frontier nodes, simultaneously predicting traversability, visual frontiers, and object similarity to enable real-time onboard semantic navigation.
EquiContact proposes a hierarchical framework consisting of a high-level vision planner, Diff-EDF, and a low-level compliant visuomotor policy, G-CompACT, for learning contact-rich manipulation policies that generalize across task configuration spaces.
The authors propose PGGT (Primitive-Grounded Geometry Transformer), a network that takes a single query image and a set of planar map primitives as input and directly predicts a 6-DoF camera pose, enabling feed-forward camera localization. The method sets new state-of-the-art localization performance on both dense surface maps and minimal layout models. Code and models are publicly available, and the paper was accepted to NeurIPS 2026.
DrawSplat proposes an Atlanta World-guided implicit Gaussian SLAM framework for indoor reconstruction. The method first estimates camera poses, sparse geometry, and gravity direction through joint point-line tracking, then infers structural planes from tracked lines using the shared gravity component of the Atlanta World prior and converts them into differentiable regularization terms on Gaussian centers and normals. The framework also combines an anchor-based implicit neural Gaussian representation with adaptive structural sampling, achieving robust tracking, high-fidelity rendering, and geometrically consistent reconstruction across multiple datasets.
AquaProbe proposes an underwater image adaptation method for frozen vision models: it measures how domain differences are distributed across encoder depths using paired clean images and physically rendered images, then allocates a fixed LoRA rank budget according to that distribution.
The study proposes a generative measurement model for encoders and derives the maximum a posteriori estimation expression and the Jacobians required for optimization, building a tightly coupled visual-inertial-encoder RGB-D SLAM system. On a self-built dataset, the system's real-time state estimation achieves a root mean square error of about 2-7 cm, used to compensate for the failure of pure vision solutions during temporary feature point loss and fast motion. The source code and the dataset containing encoder information have been made public.
This paper proposes a flexible sensing manipulation framework that directly connects the structural dynamics of flexible links with sensing, kinematics, and control. The method first identifies the dominant vibration modes of each flexible link according to response convergence and excitation bandwidth criteria. Under the condition of full-rank modal observability, m retained modes require only m independent slope measurements and one reference IMU, and the continuous link deformation is reconstructed accordingly for flexible sensing forward kinematics and a PDE-based controller.
The arXiv paper "When the Past Shapes Action: Understanding Embodied Memory" defines embodied memory as the retention of decision-relevant information beyond the current observation, and organizes existing methods around six fundamental memory requirements and eight architectural categories.
OpenVIO is a visual-inertial odometry (VIO) system released as open source under the MIT license. It supports pinhole and fisheye cameras and extends to stitched 360° images. The authors describe the design considerations for each module of the system and benchmark it against other open-source VIO systems on public datasets. The code is open source, and the paper is 7 pages with 6 figures.
This paper proposes a voxel mapping system that uses parallel computing to simultaneously generate occupancy grid maps (OGMs) and Euclidean distance transforms (EDTs), and incrementally builds a global EDT to output high-frequency local distance information for online motion planning. The system uses GPU programming to compute EDT in parallel within local volumes, then continuously integrates new observations into the global EDT via a parallel wavefront algorithm while preserving historical observations.
SAFARI treats object detection as a native channel of a unified probabilistic map, letting geometric and semantic evidence jointly drive exploration and planning. The authors completed 11 missions on a quadruped platform across three field sites, achieving an average coverage of 98.7% of the accessible mission area, with 97.6% of mission time requiring no operator intervention, and detecting objects with characteristic dimensions as small as 15 cm. Ablation experiments show that object-size awareness and safety-aware path planning are critical for hazardous-object search.
CalibSwitch is an observation-driven camera-LiDAR extrinsic calibration framework that separates absolute extrinsic recovery from selective state maintenance.
The research team proposes TRACE, an execution-time target-state risk gate that intervenes selectively while keeping the learned policy unchanged. They first used PPO to update only the planning head of a 0.6B VLA tracker; success and following rates improved across all three EVT-Bench settings, but the collision rate rose in distracted tracking scenarios.
The authors propose CONSORT, a multi-robot planning method for cooperative localization in GPS-denied environments. By optimizing only over mean-state trajectories and propagating covariance only for terminal constraint evaluation, it turns the belief-space problem into a solvable form.
Dong Hao's team at Peking University has published the paper "One Scan is Enough," which uses a LiDAR-equipped iPhone to scan a target environment once, automatically generates navigation tasks, and uses the reconstructed scene geometry for reinforcement post-training, without needing manual teleoperation demonstrations.
Why it matters: A single iPhone scan replaces manual demonstrations; 200 real-robot tests give a success-rate comparison from 25% to 85%, showing how post-training costs could fall.
Xingyan Intelligence has launched Auriga, a general embodied Harness framework. Under a zero-shot protocol that involves no task-specific pre-tuning, no reliance on expert demonstrations, single-attempt execution, and no environment resets, it achieved a 57.74% average task success rate and an average process score of 64.00 across 42 tasks and 6,300 test episodes on RoboDojo, using GPT-6 Astra as the base model. It leads the leaderboard on both metrics.
Xinzhiyuan reported on the SpatialScout benchmark, which gives a camera to an AI and places it in previously unseen virtual rooms, where the model autonomously decides when to turn, move, and capture images, and submits the 3D coordinates of target objects. The formal evaluation selects 80 indoor scenes and 3,140 target instances from iTHOR, RoboTHOR, ArchitecTHOR, ReplicaCAD, and ProcTHOR, with at most 100 interaction rounds per task; scoring considers both mean IoU and the proportion of IoU above 0%, 50%, and 90%.
On November 1, the bipedal robot "Hangxing No. 2," developed by DEEP Robotics in collaboration with Hangzhou public security traffic control authorities, made its debut at the 2026 Hangzhou Marathon, stationed at 4 key points around the Huanglong Sports Center to handle tasks such as intersection monitoring, autonomous patrols, dissuading traffic violations, and answering public inquiries.
The author uses EvergreenSLAM, a self-developed 2D lidar SLAM project, as Claude Code's spatial memory. Through the egs command line, each location corresponds to a folder and notes.md on the lidar map, with the agent writing notes itself, and a two-day real-world demonstration was completed on an XLeRobot.
The ICLR 2026 paper SAGE-3D proposes decoupling and assembling 3DGS rendering, source-mesh collision bodies, and manual semantic annotations into an executable navigation environment, modeled as a semantics- and physics-enhanced POMDP. Its object-level semantic anchoring provides category, instance ID, and bounding box, while physics-aware execution assembles collision bodies extracted from source meshes via CoACD convex decomposition and embeds them into the 3DGS scene.
KAIST AI and Holiday Robotics propose a robot-centric pointmap that transforms RGB-D observations into the robot coordinate frame while preserving the H×W grid, and adds it element-wise to RGB tokens to inject it into a pretrained VLA.
The author completed Fetch, a robotics STEM project for beginner and intermediate learners costing about 300 USD. It uses a binary-classification object detection model trained with Edge Impulse and is the author's first autonomous robot.
A team from Tianjin University and Hefei University of Technology proposes FAME, which uses Conditional Flow Matching to generate multiple exploration anchors simultaneously, then determines exploration routes through A* path planning, path clustering, and hierarchical selection. The paper was accepted to NeurIPS 2026 and selected as a Spotlight.
Reddit users discuss a paper on a tactile sensor based on mechanochromism, whose authors claim the sensor can capture pressure at fingerprint-level resolution and generate color pressure maps. The poster questions whether robots really need such high tactile resolution, and admits they are not a materials specialist and are unsure how the material actually works.
Xuanji Dynamics' Hypertron series quadruped robots won the "High-Rise Fire Rescue Track" championship and first prize in the "Load-Bearing Climbing and Material Transport" category at the 2026 Guangzhou Safety Emergency Rescue Robot Competition. The Hypertron-SW01 is equipped with a self-developed high-torque joint motor module and a multimodal fusion perception system, features IP67 protection, and continued stable motion control, perception, and communication during a 3-minute continuous bake test at 120°C–130°C.
A University of Hong Kong team has open-sourced GenSfM, which uses a native 3D diffusion model to jointly denoise object shape and per-view camera poses within a single multimodal sequence, requiring only 2 to 4 uncalibrated images and no camera intrinsics or extrinsics. Evaluated on Toys4k and GSO, it improves novel-view PSNR by 3.2–4.5 dB over the multi-view baseline TRELLIS-M, reduces median Chamfer distance by about 64%–74%, and achieves pose Acc@5° above 96%.
Why it matters: The work turns a native 3D diffusion model into the main inference mechanism for joint denoising of shape and camera poses, and releases code and models as a reference for sparse-view reconstruction.
REEMAN customized the forks on its Titan Mini autonomous forklift for a high-end electronic materials factory in Anhui, enabling it to directly handle the customer's existing wheeled mobile material tank trolleys without additional pallets. The trolley weighs about 275 kg when loaded, the Titan Mini has a rated payload of up to 2,000 kg, the one-way point-to-point transport distance is about 260 m, and it uses LiDAR SLAM navigation; under the current site configuration, a full work cycle takes about 4.5 minutes.
A weekend PoC mounts an old Redmi 9 on a Thymio II using Lego, with the phone running ML Kit face detection (including smile probability) and image annotation, rough lighting/motion estimation, animated expressions, and voice output, while the robot only executes wheel-speed targets and sends back sensor data. Behavior is a small state machine: cruise → back up → turn → observe → stop steadily; when a face is detected it enters a "meet" state and can answer "what do you see." Obstacle-avoidance learning is simple: after a collision it tries turning for 1.2s, 2.4s, or 3.8s, scored by the free-driving time before the next contact.
REEMAN demonstrated a library inventory robot built on the Hot Wheels robot chassis, completing a mobility demo in indoor bookshelf aisles, but did not provide OCR recognition accuracy or inventory results. The chassis uses Laser SLAM and V-SLAM for mapping and localization, combined with lidar, 3D vision, and multi-sensor fusion for obstacle avoidance; published specifications include a 270° laser scanning angle and a maximum detection range of 25 meters.
The inaugural Global Ocean Robotics Day was held on October 21, with Constructor University as one of 10 in-person hubs worldwide. The event focused on how artificial intelligence and autonomous robots are transforming ocean research and exploration. Topics covered marine AI, autonomous underwater vehicles, and ocean observation robotic systems, including mapping ocean currents using marine vehicle trajectories, simulating autonomous docking with digital twins, and underwater robot swarms. The university's Ocean Systems and Robotics research group also demonstrated several autonomous marine robots developed by the team on site.
Maximo Robotics announced that its US-made robotic systems have completed more than 300,000 construction tasks under task-based robotics-as-a-service contracts, and that workers collaborating with the Max 4.0 system achieved installation efficiency 2.5 times that of manual labor per person per shift hour.
Taiying Technology Group has built an embodied AI robot mountain-climbing training ground at the foot of Mount Tai, replicating the 3.8-kilometer, more than 3,100-step terrain from Mount Tai's Hongmen Gate to Zhongtianmen, with 11 test tracks for humanoid robots and quadruped robot dogs preparing for the first CMG World Robot Mount Tai Climbing Competition. The training ground includes special tracks such as low-recognition steps, steep, slippery, narrow steps, and S-shaped sharp-turn steps, focusing on assessing environmental perception, dynamic balance, and gait-switching capabilities. It has been selected as one of Shandong Province's first batch of key-supported regional hub training grounds.
Smart wheelchair makers are redefining their products as 'travel robots' or home entry points for embodied AI, seeking to attract younger users with 'light autonomous driving.' Strutt's EV1 carries 20 sensors—2 LiDARs, 10 ToF sensors, 6 ultrasonic radars, and 2 cameras—enabling 360° environmental perception and 3D mapping; Xcan Mate X from Xingcan Intelligence uses a LiDAR, ultrasonic radar, and camera setup. By 2026, the sector has seen at least 9 publicly disclosed funding rounds, but buyers remain concentrated among people aged 30-50 purchasing for their parents.
An engineering team at the University of Pennsylvania has proposed Fast Feature Field (F³), which converts the tens of millions of signals per second from event cameras into a general-purpose representation usable by existing computer vision software. The paper was published in Science Robotics.
The author debugged and fine-tuned a motion controller based on the MPC/WBC framework for a self-developed quadruped robot, and used depth-camera images to build height maps. The robot can now climb stairs more reliably, step over obstacles up to 11 cm high (with 18 cm ground clearance), and handle soft ground with greater confidence. The controller is written in C++, is ROS2-compatible, and has been open-sourced in the voltdog/mors_quadruped repository on GitHub. Next, the author will focus on improving software reliability, including automated device connection tests and standing up from any posture.
Xuanyuan has extended its explosion-proof inspection robots, already delivered at scale to chemical enterprises such as Wanhua Chemical, to urban gas pipeline networks; the solution has been put into operation at a gas station. The robot first scans and builds a model to calibrate instrument and valve points, then automatically inspects on a periodic basis via the backend, integrating image-based meter reading, infrared thermal imaging for temperature measurement, and gas-sensor leak detection; its four-wheel, eight-drive omnidirectional chassis can pass through narrow passages and recharge autonomously. On-site data is transmitted back to the central control in real time, anomalies are flagged by the system and reviewed by on-duty personnel, and manual steps remain in risk handling.